This invention generally relates to a novel process for making a carboxylic acid salt. More particularly, this invention relates to a process for dehydrogenating a primary alcohol (especially an amino alcohol, such as diethanolamine) to make a carboxylic acid salt (such as disodium iminodiacetic acid) using a copper-containing or silver-containing catalyst which also contains other metals that provide desirable characteristics, such as durability. This invention also generally relates to novel copper-containing and silver-containing catalysts that may be used in such a process, and to processes for making such catalysts.
Carboxylic acid salts are useful in various applications. For example, salts of iminodiacetic acid may be phosphonomethylated to form N-(phosphonomethyl)iminodiacetic acid (xe2x80x9cPMIDAxe2x80x9d), which, in turn, may be oxidized to form N-(phosphonomethyl)glycine (known in the agricultural chemical industry as xe2x80x9cglyphosatexe2x80x9d). See, e.g., Gentilcore, U.S. Pat. No. 4,775,498 (disclosing a method to phosphonomethylate a salt of iminodiacetic acid); Ebner, et al., PCT/US99/03402 (disclosing a method for oxidizing PMIDA). Salts of nitrilotriacetic acid, for example, are excellent chelating agents, and consequently may be used as detergent builders, water-softening agents, scouring aids, dyeing assistants, paper-coating agents, scale inhibitors, and agents for preventing soap degeneration. And many carboxylic acid salts (e.g., salts of glycine, salts of iminodiacetic acid, etc.) may also be neutralized to their corresponding acids and then used, for example, as chelating agents; in food preparations; and as raw materials for making pharmaceuticals, agricultural chemicals, and pesticides. See, e.g., Franz, et al., Glyphosate: A Unique Global Herbicide (ACS Monograph 189, 1997) at pp. 234-41 (disclosing the use of glycine and iminodiacetic acid compounds as raw materials to form N-(phosphonomethyl)glycine).
It has long been known that a carboxylic acid salt may be prepared from a primary alcohol by dehydrogenating the alcohol using a copper-containing or silver-containing catalyst. In 1945, Chitwood first reported forming a carboxylic acid salt (specifically, the potassium salt of glycine) by oxidizing a primary alcohol (specifically, monoethanolamine) in an alkaline environment (specifically, in a mixture containing potassium hydroxide) using a copper-containing catalyst (specifically, copper metal or cupric oxide, which reportedly was reduced to copper metal under the reaction conditions) or a silver-containing catalyst (specifically, silver metal or silver oxide, which reportedly was reduced to silver metal under the reaction conditions). See Chitwood, U.S. Pat. No. 2,384,817. Chitwood, however, reported that copper-containing compounds are disadvantageous for this reaction because the copper coagulates over time, thereby causing the copper-containing compounds to have a short duration of maximum catalytic activity. Chitwood also reported that silver-containing compounds have relatively low activity (the silver oxide also reportedly coagulates over time).
In 1988, Goto et al. reported forming a carboxylic acid salt by oxidizing an ethanolamine compound in an alkaline solution (specifically, an aqueous solution containing the hydroxide of an alkali metal or an alkaline earth metal) using Raney copper. See Goto et al., U.S. Pat. No. 4,782,183. Goto et al. reported selectivities of at least 94.8% when dehydrogenating monoethanolamine, diethanolamine, and triethanolamine to form salts of glycine, iminodiacetic acid, and nitrilotriacetic acid, respectively. Raney copper, however, is disadvantageous because (like Chitwood""s copper-containing compounds) Raney copper deactivates over time. See, e.g., Franczyk, U.S. Pat. No. 5,292,936, Table 1 (showing the reaction time for Raney copper to increase from 4 to 8 hours over 9 cycles).
Various developments have been reported which address the instability of copper-containing catalysts when used to dehydrogenate primary alcohols. Although these developments have made the use of copper catalysts more commercially viable, their results are still not entirely satisfactory.
Franczyk, for example, reports that copper-containing catalysts particularly Raney copper) can be stabilized by using such a catalyst which also contains 50 to 10,000 parts per million of one or more various other metals selected from the group consisting of chromium, titanium, niobium, tantalum, zirconium, vanadium, molybdenum, tungsten, cobalt, nickel, bismuth, tin, antimony, lead, and germanium, with vanadium, chromium, and molybdenum being the more preferred metals. See Franczyk, U.S. Pat. Nos. 5,292,936; 5,367,112; and 5,739,390. Although such metals do tend to impart a stabilizing effect to a copper catalyst, this effect often decreases over time. See, e.g., Franczyk patents, Table 2 (showing the reaction time decreasing from 5.8 hours to 8.0 hours over 25 cycles) and Table 4 (showing the reaction time decreasing 3.1 to 5.5 hours over 12 cycles). This decrease is due, at least in part, to the fact that such metals tend to leach over time as the catalyst is used, particularly where the primary alcohol or the dehydrogenation product is a chelating agent (e.g., a salt of iminodiacetic acid).
Ebner et al. report using a catalyst comprising copper supported on an alkali-resistant support (particularly a carbon support) to dehydrognate primary alcohols to make carboxylic acid salts. See Ebner et al., U.S. Pat. No. 5,627,125. This catalyst also comprises about 0.05 to about 10% by weight of a noble metal to anchor and disperse the copper to the support. Although Ebner et al. report shorter reaction times with their catalyst relative to previously disclosed copper-containing catalysts, their catalyst is costly due to the need for the noble metal to anchor the copper to the support. In addition, the added volume of Ebner et al.""s catalyst due to the carbon support can, in some instances, make handling the catalyst cumbersome, consequently reducing throughput. Further, Ebner et al.""s catalyst often loses activity over time with use (although the rate of deactivation is often less than the rate of deactivation of the Franczyk catalysts). See, eg., Ebner et al., Table 1 (showing the reaction time increasing from 103 to 150 minutes over 9 cycles) and Table 2 (showing the reaction time increasing from 61 to 155 minutes over 8 cycles). As with the Franczyk catalysts, this problem tends to arise particularly where the primary alcohol or the dehydrogenation salt product is a chelating agent.
Other reported copper-containing catalysts contain a non-carbon support, such as, SiO2, Al2O3, TiO2, ZrO2, and the like. See, eg., Akzo Nobel, WO 98/13140 (disclosing a catalyst consisting of copper on ZrO2). These supports, however, tend to be vulnerable to attrition under the reaction conditions normally present when dehydrogenating a primary alcohol, and are therefore usually less suitable than Ebner et al.""s carbon supports. This vulnerability to attrition tends to also cause these supports to exhibit poor filtration characteristics.
Use of copper-containing and silver-containing catalysts in other types of oxidation reactions has also been reported. Applicants, however, are unaware of any such disclosures which address the problems associated with copper-containing or silver-containing catalysts in processes involving the dehydrogenation of primary alcohols to form carboxylic acid salts.
Bournonville et al. report forming a ketone by dehydrogenating a secondary alcohol using a Raney nickel catalyst containing 0.1 to 10% by weight of copper, silver, gold, tin, lead, zinc, cadmium, indium, or germanium. See Bournonville et al., U.S. Pat. No. 4,380,673. This reaction, however, does not form a carboxylic acid saltxe2x80x94forming a carboxylic acid salt would further require the cleavage of an alkyl group from the carbonyl group and the subsequent attachment of a hydroxy salt to the carbonyl group. In addition, Bournonville et al. report that their reaction is catalyzed by the Raney nickel, and that the function of the additional metal (e.g., copper or silver) is to suppress hydrogenolysis side reactions. See Bournonville et al., col. 3, lines 45-47. This is in contrast to dehydrogenation reactions of primary alcohols using copper catalysts, such as Raney copper, where catalytic activity is provided primarily by copper atoms near the surface of the catalyst.
Yamachika et al. report forming benzaldehydes by reducing benzonitriles in the presence of acid and a Raney nickel catalyst which has been pre-treated with a copper salt solution. See Yamachika et al., U.S. Pat. No. 4,500,721. Yamachika et al. disclose that the conditions of catalyst pre-treatment should be sufficient to form a catalyst which contains 5 to 80% (more preferably 10 to 60%) by weight of copper. Yamachika et al. report that the presence of the copper increases the yield of benzaldehydes during the reaction. This reaction, however, is conducted in an acidic environment, is not directed to dehydrogenating primary alcohols (or any other alcohols), and does not form carboxylic acid salts.
Thus, although positive advances have been reported for converting a primary alcohol to a carboxylic acid salt using a copper-containing catalyst, there continues to be a need for a more economical liquid-phase process which uses a catalyst that has high surface area, has high activity, and exhibits stability (i.e., maintains its activity) over time with usage. This need particularly exists where the primary alcohol substrate and/or carboxylic acid salt product is a chelating agent (e.g., a salt of iminodiacetic acid).
This invention provides for a novel and improved liquid-phase process for dehydrogenating primary alcohols to form salts of carboxylic acids. In particular, this invention provides for a dehydrogenation process that can use an economically advantageous catalyst (e.g., a catalyst that does not require the presence of expensive precious metals). This invention also provides for a dehydrogenation process that uses a catalyst that has a high surface area (e.g., at least about 20 m2/g, and more typically at least about 35 m2/g). This invention additionally provides for a dehydrogenation process that uses a catalyst that maintains its activity, even in a mechanically-stirred, alkaline liquid containing one or more chelating agents (i.e., the reaction conditions where copper catalyst deactivation has traditionally been most pronounced). This invention further provides for a dehydrogenation process that can use a copper-containing catalyst which has less volume per unit surface area of copper than the traditional catalysts containing copper supported on carbon.
Briefly, therefore, this invention is directed to a process for making a salt of a carboxylic acid. This process comprises contacting a catalyst with an alkaline mixture comprising a primary alcohol.
In one embodiment, the catalyst comprises a metal support (preferably a metal sponge support) coated with copper. The support comprises at least about 10% (by weight) non-copper metal. The copper-containing coating comprises from about 0.005 to about 0.5 grams of copper (per gram of said metal support).
In another embodiment, the catalyst comprises at least about 15% (by weight) non-copper metal, and at least about 10% (by weight) copper. In a particularly preferred embodiment, this catalyst is in the form of a metal sponge. In another particularly preferred embodiment, the catalyst comprises less than about 1% (by weight) metal oxide. In yet another particularly preferred embodiment, the catalyst comprises greater than about 1% (by weight) nickel, tin, chromium, tungsten, titanium, niobium, tantalum, vanadium, molybdenum, manganese, bismuth, antimony, lead, germanium, or a combination thereof.
In another embodiment, the catalyst comprises a metal support (preferably a metal sponge support) coated with silver. The support comprises at least about 10% (by weight) non-silver metal. The silver-containing coating comprises from about 0.005 to about 0.5 grams of silver (per gram of said metal support).
In another embodiment, the catalyst comprises at least about 15% (by weight) non-silver metal, and at least about 10% (by weight) silver. In a particularly preferred embodiment, this catalyst is in the form of a metal sponge.
In another embodiment, the catalyst comprises (i) a metal sponge; (ii) at least about 70% (by weight) metal, and less than about 1% (by weight) metal oxide; or (iii) at least about 70% (by weight) metal, and greater than about 1% (by weight) nickel, tin, chromium, tungsten, titanium, niobium, tantalum, vanadium, molybdenum, manganese, bismuth, antimony, lead, germanium, or a combination thereof. In this embodiment, the catalyst may be identified in that a reference consumption of at least 75% of a diethanolamine substrate may be achieved within a time period of about 3 hours under constant maximum pressure when said catalyst is contacted with an alkaline mixture containing said substrate to form disodium iminodiacetic acid and H2 under the following reference conditions: (i) the alkaline mixture initially consists of 0.36 moles of diethanolamine, 0.77 moles of NaOH, and 70 grams of water; (ii) the weight of catalyst contacted with the alkaline mixture is equal to 5% of the weight of the alkaline mixture; (iii) the diethanolamine dehydrogenation is conducted in a reactor having a head space of no greater than 4 liters, and initially containing a N2 atmosphere at atmospheric pressure; (iv) both the catalyst and the alkaline mixture are at 150xc2x0 C. when contacted, and maintained at 150xc2x0 C. during said diethanolamine dehydrogenation; and (v) the pressure in the reactor is allowed to rise autogenously from atmospheric pressure at the beginning of the reaction to a maximum constant pressure of 135 psig, after which the reactor is continuously vented to maintain said maximum constant pressure of 135 psig. Here, the xe2x80x9ctime period under constant maximum pressurexe2x80x9d is the period between the time at which the pressure first reaches 135 psig and the subsequent time at which the evolution of H2 from the reaction has first declined to 5 sccm. Also, the xe2x80x9creference consumptionxe2x80x9d of diethanolamine substrate is the total consumption as measured at the end of said maximum constant pressure time period.
This invention also provides for a novel and improved copper-containing catalyst which may, for example, be used in liquid-phase oxidation reactions, particularly liquid-phase dehydrogenation reactions which convert primary alcohols to carboxylic acid salts. More specifically, this invention provides for a copper-containing catalyst that is economically advantageous because, for example, it does not require the presence of expensive precious metals. This invention also provides for a catalyst that has a high surface area. This invention additionally provides for a catalyst that maintains its activity with use over time, even in a mechanically-stirred, alkaline liquid containing one or more chelating agents. This invention further provides for a copper-containing catalyst that has less volume per unit of copper surface area than the traditional catalysts comprising copper supported on carbon, thereby providing greater filterability. This invention still further provides for a copper-containing catalyst that has greater resistance to attrition than traditional catalysts comprising copper or comprising copper on carbon supports.
Briefly, therefore, this invention is directed to a copper-containing oxidation catalyst (the term xe2x80x9coxidationxe2x80x9d includes, but is not limited to, dehydrogenation reactions). This catalyst comprises a metal support (preferably a metal sponge support) coated with copper. The support comprises at least about 10% (by weight) non-copper metal and at least 10% (by weight) copper. The copper-containing coating comprises from about 0.005 to about 0.5 grams of copper (per gram of said metal support).
This invention also is directed to a process for making such a copper-containing catalyst. This process comprises depositing a copper-containing coating onto a surface of a metal support where (a) the metal support comprises at least about 10% (by weight) copper, and at least about 10% (by weight) non-copper metal; and (b) the copper-containing coating comprises from about 0.005 to about 0.5 grams of copper (per gram of said metal support).
This invention further provides for a novel silver-containing catalyst which may, for example, be used in liquid-phase oxidation reactions, particularly liquid-phase dehydrogenation reactions which convert primary alcohols to carboxylic acid salts. More specifically, this invention also provides for a silver-containing catalyst that has a high surface area. This invention additionally provides for a silver-containing catalyst that maintains its activity with use over time, even in a mechanically-stirred, alkaline liquid containing one or more chelating agents.
Briefly, therefore, this invention is directed to a silver-containing oxidation catalyst. This catalyst comprises a metal support (preferably a metal sponge support) coated with silver. The support comprises at least about 10% (by weight) silver and at least about 10% (by weight) non-silver metal. The silver-containing coating comprises from about 0.005 to about 0.5 grams of silver (per gram of said metal support).
This invention is also directed to a process for making such a silver-containing catalyst. This process comprises depositing a silver-containing coating onto a surface of a metal support where (a) the metal support comprises at least about 10% (by weight) silver, and at least about 10% (by weight) non-silver metal; and (b) the silver-containing coating comprises from about 0.005 to about 0.5 grams of silver (per gram of said metal support).
Other objects and features of this invention will be in part apparent and in part pointed out hereinafter.
The process of this invention may generally be used to convert any primary alcohol to a carboxylic acid salt. As used herein, a xe2x80x9cprimary alcoholxe2x80x9d is any alcohol comprising a hydroxy group attached to a carbon which is bound to two hydrogen atoms, i.e., Rxe2x80x94CH2OH.
This process dehydrogenates a primary alcohol to yield both a carboxylic acid salt and hydrogen gas. Typically, this reaction is carried out in a heated reaction zone containing the primary alcohol, a base, and a copper-containing or silver-containing catalyst. An example of this reaction is the dehydrogenation of monoethanolamine in a heated reaction zone containing KOH to form hydrogen gas and the potassium salt of glycine: 
Another example of this reaction is the dehydrogenation of diethanolamine (sometimes described in the art as xe2x80x9cDEAxe2x80x9d) in a heated reaction zone containing NaOH to form hydrogen gas and disodium iminodiacetic acid (sometimes described in the art as xe2x80x9cDSIDAxe2x80x9d): 
An additional example is the dehydrogenation of an N-alkyl-monoethanolamine to form a salt of an N-alkyl-glycine. The alkyl group could be, for example, methyl (xe2x80x94CH3). In that instance, the dehydrogenation product would be a salt of N-methyl-glycine (i.e., a salt of sarcosine): 
A further example is the dehydrogenation of triethanolamine to form a salt of nitrilotriacetic acid: 
A. Preferred Primary Alcohol Substrate
This process is particularly useful for primary alcohols which contain amino groups or other functionalities which are reactive and susceptible to side reactions. In particular, xcex2-amino alcohols are susceptible to dehydrogenation of the Cxe2x80x94N bond and subsequent dealkylation, consequently leading to the formation of typically undesirable side products.
In one embodiment of this invention, the primary alcohol is an alkanolamine (i.e., a compound wherein the nitrogen of an amine functionality is bonded directly to the carbon of an alkyl alcohol). In this embodiment, the primary alcohol preferably has formula (I): 
wherein n is an integer ranging from 2 to 20; and R1 and R2 are independently hydrogen, hydrocarbyl, or substituted hydrocarbyl.
A hydrocarbyl may be any group consisting exclusively of carbon and hydrogen. The hydrocarbyl may be branched or unbranched, may be saturated or unsaturated, and may comprise one or more rings. Suitable hydrocarbyl groups include alkyl, alkenyl, alkynyl, and aryl groups. They also include alkyl, alkenyl, alkynyl, and aryl groups substituted with other aliphatic or cyclic hydrocarbyl groups, such as alkaryl, alkenaryl, and alkynaryl.
A substituted hydrocarbyl may be any hydrocarbyl wherein at least one hydrogen atom has been substituted with an atom other than hydrogen or a group of atoms containing at least one atom other than hydrogen. For example, the hydrogen atom may be substituted with a halogen atom, such as a chlorine or fluorine atom. The hydrogen atom alternatively may be substituted with an oxygen atom or a group containing an oxygen atom to form, for example, a hydroxy group, an ether, an ester, an anhydride, an aldehyde, a ketone, or a carboxylic acid. The hydrogen atom also may be replaced with a group containing a nitrogen atom to form, for example, an amide or a nitro group. In addition, the hydrogen atom may be substituted with a group containing a sulfur atom to form, for example, xe2x80x94SO3H.
Typically, R1 and R2 are independently either: hydrogen; xe2x80x94(CH2)xxe2x80x94CH3, x being an integer ranging from 0 to about 19 (particularly from 1 to 6, and even more particularly 1); xe2x80x94(CH2)yxe2x80x94OH, y being an integer ranging from 1 to about 20 (especially from 2 to 6); (CH2)zxe2x80x94COOH, z being an integer ranging from 1 to about 19 (especially from 1 to 5); or phosphonomethyl.
In some preferred embodiments, R1 and R2 are both hydrogen (i.e., the amine functionality shown in formula (I) is a primary amine). An example of such an alcohol is monoethanolamine.
In other preferred embodiments, R1 is hydrogen and R2 is hydrocarbyl or substituted hydrocarbyl (i.e., the amine functionality shown in formula (I) is a secondary amine). Examples of primary alcohols in which R2 is hydrocarbyl include N-methylethanolamine, N-ethylethanolamine, N-isopropylethanolamine, N-butylethanolamine, and N-nonylethanolamine. Examples of primary alcohols in which R2 is a substituted hydrocarbyl include primary alcohols wherein R2 is xe2x80x94(CH2)yxe2x80x94OH and y is an integer ranging from 1 to about 20 (more preferably from 1 to 6). An example of such an alcohol is diethanolamine. Other examples of primary alcohols wherein R2 is a substituted hydrocarbyl include N-(2-aminoethyl)ethanolamine, N-(3-aminopropyl)ethanolamine, N-(carboxymethyl)ethanolamine, and N-(phosphonomethyl)ethanolamine. N-substituted ethanolamines, for example, may be prepared using the various methods known in the art. For example, a ketone may be condensed with monoethanolamine in the presence of H2, a solvent, and a noble metal catalyst. This reaction is described in, for example, Cope, A. C. and Hancock, E. M. J. Am. Chem. Soc., 64, 1503-6 (1942). N-substituted ethanolamines also may be prepared by combining a mono-substituted amine (such as methylamine) and ethylene oxide to form the mono-substituted ethanolamine. This reaction is described by, for example, Y. Yoshida in Japanese Patent Application No. 95-141575.
In yet other preferred embodiments, both R1 and R2 are independently hydrocarbyl or substituted hydrocarbyl (i.e., the amine functionality shown in formula (I) is a tertiary amine). Examples of primary alcohols in which R1 and R2 are independently hydrocarbyl include N,N-dimethylethanolamine, N,N-diethylethanolamine, and N,N-dibutylethanolamine. Examples of primary alcohols in which R1 is hydrocarbyl and R2 is substituted hydrocarbyl include primary alcohols wherein R2 is xe2x80x94(CH2)yxe2x80x94OH and y is an integer ranging from 1 to about 20 (more preferably from 1 to 6). Such alcohols include, for example, N-methyldiethanolamine, N-ethyldiethanolamine, N-isopropyldiethanolamine, and N-butyldiethanolamine. Other examples of primary alcohols in which R1 is hydrocarbyl and R2 is substituted hydrocarbyl include N-ethyl, N-(2-aminoethyl)ethanolamine; N-ethyl, N-(2-aminoethyl)ethanolamine; and N-methyl, N-(3-aminopropyl)ethanolamine. Examples of primary alcohols in which R1 and R2 are independently substituted hydrocarbyl include primary alcohols wherein R1 and R2 are independently xe2x80x94(CH2)yxe2x80x94OH and y is an integer ranging from 1 to about 20 (more preferably from 1 to 6). An example of such an alcohols is triethanolamine. Other examples of primary alcohols in which R1 and R2 are independently substituted hydrocarbyl include tetra(2-hydroxyethyl)ethylenediamine and N-(phosphonomethyl),N-(carboxymethyl)ethanolamine.
B. Catalyst
It has been found in accordance with this invention that the softness of copper is at least one of the reasons that many traditional copper-containing catalysts (particularly copper sponge catalysts, such as those described by Goto et al. in U.S. Pat. No. 4,782,183) deactivate over time. More specifically, as such catalysts are used, their surfaces tend to lose surface area and the catalyst particles themselves tend to agglomerate (this agglomeration, in turn, reduces access by the reactants to the catalyst""s active sites). These effects are particularly pronounced when the traditional catalysts are used in a stirred-tank reactor (or otherwise subjected to mechanical agitation). Both the loss of surface area and the agglomeration of the catalyst particles reduce the surface area of the catalyst, thereby reducing activity of the catalyst. It has been discovered in accordance with this invention, however, that the rate of deactivation can be significantly reduced by combining the copper with at least one other metal which, at least in part, provides strengthening characteristics to the copper to make a more durable catalyst.
Because silver is a relatively soft metal like copper, the same principles tend to apply to silver-containing catalysts. Silver-containing catalysts, however, are generally less preferred due to their relatively greater cost compared to copper-containing catalysts. Thus, most of the following discussion will focus on copper-containing catalysts. Nevertheless, it should be recognized that this discussion generally applies to silver-containing catalysts as well.
1. Catalysts Comprising Copper on a Metal Support
In one embodiment of this invention, the catalyst comprises a metal support having a copper coating on its surface. This coating preferably comprises from about 0.005 to about 0.5 grams (more preferably from about 0.02 to about 0.3 grams, even more preferably from about 0.05 to about 0.2 grams, still even more preferably from about 0.08 to about 0.15 grams) of copper (per gram of said metal support). In other words, for every gram of metal support that the catalyst contains, the catalyst also preferably contains a copper-containing coating which comprises from about 0.005 to about 0.5 grams (more preferably from about 0.02 to about 0.3 grams, even more preferably from about 0.05 to about 0.2 grams, still even more preferably from about 0.08 to about 0.15 grams) of copper.
a. The Metal Support
The metal support may comprise a wide variety of compositions. In general, however, at least about 10% (by weight) of the metal support is non-copper metal. In one particularly preferred embodiment, at least 65% (more preferably at least about 80%, even more preferably at least about 85%, and still even more preferably at least about 90%) by weight of the metal support is non-copper metal (this non-copper metal may comprise a single metal or multiple metals). In another particularly preferred embodiment, on the other hand, at least 50% (more preferably from about 60 to about 80%) by weight of the metal support is copper.
The metal or alloy from which the metal support is made preferably has a tensile strength and/or yield strength which is greater than copper alone. It is particularly preferred for the composition to have a yield strength of greater than about 70 Mpa, more preferably greater than 100 Mpa, and even more preferably at least 110 Mpa. It is also particularly preferred for the composition to have a tensile strength of greater than 221 Mpa, more preferably greater than 275 Mpa, and even more preferably greater than 300 Mpa. For example, a composition containing 70% (by weight) copper and 30% (by weight) zinc reportedly has a yield strength of 124 Mpa and a tensile strength of 331 Mpa; a composition containing 90% (by weight) copper and 10% (by weight) nickel reportedly has a yield strength of 110 Mpa and a tensile strength of 303 Mpa; and a composition containing 70% (by weight) copper and 30% (by weight) nickel reportedly has a yield strength of 138 Mpa and a tensile strength of 372 Mpa. See A. S. Krisher and O. W. Siebert in Perry""s Chemical Engineers"" Handbook, pp. 23-42 to 23-49 (6th ed., R. H. Perry, D. Green, and J. O. Maloney, eds, McGraw Hill, New York, N.Y. 1984).
In many instances, it is preferred for the non-copper metal in the support to be relatively non-reactive in the alkaline (and often chelating) environments of this process. Such relatively non-reactive metals include, for example, nickel, gold, palladium, and platinum. Of these metals, nickel is typically the more preferred because, for example: (1) nickel generally costs less than the other metals, and (2) depositing copper onto a nickel-containing support is typically less difficult relative to depositing copper onto a support containing a significant amount of the other listed metals because copper may be deposited onto a nickel-containing support using the simple process of electrochemical displacement deposition (there are, however, other techniques (e.g., electroless plating) which may often be used to coat copper onto supports comprising gold, palladium, and/or platinum).
It should be recognized that, although it is preferred to use a support comprising a metal which is relatively non-reactive in the reaction environments of this method, other metals (e.g., zinc, cobalt, iron, and tin) which show some greater degree of reactivity in alkaline and/or chelating environments also may often be suitable. This is particularly true because the copper on the surface of the metal support tends to act as a shield to protect the metal in the support from the reaction environment. It is also particularly true where the less-alkaline-resistant metals provide advantages which are not provided by the more-alkaline-resistant metals. For example, it is often desirable to deposit copper onto the surface of the metal support using electrochemical displacement deposition (also described in the art as xe2x80x9cimmersion platingxe2x80x9d). In that instance, the metal support preferably contains metal which has a reduction potential to the metal which is less than the reduction potential to the metal of copper, i.e., a reduction potential to the metal of less than about +343 mVolts vs. NHE (normal hydrogen electrode). Metals having such a reduction potential include, for example, nickel, zinc, tin, iron, and cobalt. The presence of such a metal near the surface of the support allows for simple deposition of copper metal on the surface of the support by contacting the surface with a copper salt (normally a Cu(II) salt) solution. More specifically, during displacement deposition, such a metal near the surface of the support tends to oxidize (and go into solution as an ion) when contacted with a copper ion solution. As this occurs, the copper ions in solution near the support surface are reduced to copper metal, which, in turn, forms a coating on the surface of the support. The reaction which occurs, for example, when a support comprising nickel is contacted with a copper salt solution is:
Cu2++Ni0xe2x86x92Cu0+Ni2+
It should be recognized that when coating silver onto a metal support using electrochemical displacement deposition, the metal support preferably contains metal which has a reduction potential to the metal which is less than the reduction potential to the metal of silver, i.e., a reduction potential to the metal of less than about +800 mVolts vs. NHE.
As the foregoing suggests, when the catalyst is prepared by depositing copper onto the surface of the support using displacement deposition, it is particularly preferable to use a nickel-containing support because nickel has at least three desirable characteristics: (1) an reduction potential to the metal which is less than the reduction potential to the metal of copper, (2) relative stability in the reaction conditions of this invention, and (3) greater mechanical strength and resistance to attrition than copper.
When the metal support comprises more than one metal, it is preferred that at least about 80% by weight (more preferably at least about 85% by weight, even more preferably at least about 90% by weight, and still even more preferably essentially all) of the metals in the support are in the form of an alloy. In a particularly preferred embodiment, the metals form a substitutional alloy (also known as a xe2x80x9cmonophasic alloyxe2x80x9d), wherein the alloy has a single, continuous phase. Although multiphasic alloys (i.e., alloys comprising at least 2 discrete phases) may be used, monophasic alloys are generally preferred because it is difficult to evenly distribute copper onto a multiphasic support surface because copper tends to preferentially coat the copper-rich portions relative to the copper-poor portions of the surface. Whether the alloy is monophasic or multiphasic will depend on the components of the alloy and their concentrations. Typically, for example, metal supports consisting essentially of nickel and copper are monophasic at any nickel concentration. But when, for example, the support consists essentially of copper and zinc, there are many zinc concentrations (typically, concentrations greater than about 35% by weight) which lead to the alloy being bi-phasic.
It should be recognized that the support may also comprise non-metal atoms (e.g., boron, carbon, silicon, nitrogen, phosphorus, etc.) in addition to the metal atoms. An alloy containing such non-metal is typically described in the art as an xe2x80x9cinterstitial alloy.xe2x80x9d Supports comprising such an alloy may have various advantages, such as enhanced mechanical strength. Typically, however, the catalyst comprises at least about 70% metal.
In a particularly preferred embodiment, the metal support is a metal sponge. As used herein, the term xe2x80x9cmetal spongexe2x80x9d refers to a finely divided and porous form of metal having a surface area of at least about 20 m2/g, and more typically at least about 35 m2/g (preferably from about 35 to about 100 m2/g, and even more preferably from about 80 to about 100 m2/g). Such surface area may be measured using, for example, the B. E. T. (Brunauer/Emmett/Teller) method which is well known in the art. It has been found in accordance with this invention that if copper is coated onto the surface of a metal sponge support, the resulting material exhibits the mechanical strength and high surface area of the sponge support combined with the desired catalytic activity of the copper.
Metal sponges are often generally described in the art as xe2x80x9cRaney metals.xe2x80x9d However, because W. R. Grace and Co. owns trademark rights in the word xe2x80x9cRaneyxe2x80x9d for describing metal sponges, Applicants use the term xe2x80x9cmetal spongexe2x80x9d rather than xe2x80x9cRaney metalxe2x80x9d to ensure that this patent is not limited to the use of W. R. Grace and Co.""s metal sponges.
Typically, the preferred average particle size of the metal sponge is at least about 0.1 xcexcm, preferably from about 0.5 to about 100 xcexcm, more preferably from about 15 to about 100 xcexcm, even more preferably from about 15 to about 75 xcexcm, and still even more preferably from about 20 to about 65 xcexcm.
Sponge supports can be prepared by techniques generally known to those skilled in the art. See, generally, E. Lieber and F. L. Morritz, Adv. Catal., 5, 417 (1953) (a general review directed to sponge metals). In general, techniques for making metal sponges comprise forming an alloy which contains about 50% (by weight) of a leachable metal (typically aluminum) and about 50% (by weight) of the desired metal(s); grinding the alloy to a desired particle size; and treating the alloy particles with an aqueous solution of an alkali metal hydroxide (preferably NaOH) to leach at least a portion of the leachable metal from the alloy. It is often preferred to conduct the leaching at a temperature of less than about 50xc2x0 C. (more preferably no greater than about 40xc2x0 C., and even more preferably from about 20 to about 40xc2x0 C.). As the leachable metal leaches from the particle, it leaves behind voids (e.g., pores) which dramatically increase the surface area of the particle.
It should be recognized that the above-described technique is not the only method for making sponge metals. An iron sponge, for example, may be formed by reducing iron oxide at such low temperatures that melting does not occur, typically by mixing iron oxide and coke and applying a limited increase in temperature. See Hawley""s Condensed Chemical Dictionary, 13th Ed., p. 621 (Rev. by Richard J. Lewis, Sr., Van Nostrand Reinhold, New York, N.Y. 1997).
References describing the preparation of nickel sponges include, for example, Augustine, Robert, L. Catalytic Hydrogenation Techniques and Applications in Organic Synthesis (Marcel Dekker, Inc., 1965), appendix at pp. 147-149. See also, Hawley""s Condensed Chemical Dictionary, 13th Ed., p. 955 (Rev. by Richard J. Lewis, Sr., Van Nostrand Reinhold, New York, N.Y. 1997) (describing the generally recognized technique of making sponge nickel by leaching aluminum from an alloy containing 50% by weight nickel and 50% by weight aluminum using a 25% by weight caustic soda solution).
References describing the preparation of nickel/copper sponges include, for example, D. J. Young, M. S. Wainwright, and R. B. Anderson, J. Catal., 64, 116 (1980). Such references also include, for example, M. S. Wainwright and R. B. Anderson, J. Catal., 64, 124 (1980).
References describing the preparation of copper/zinc sponges include, for example, A. J. Bridgewater, M. S. Wainwright, D. J. Young, and J. P. Orchard, Appl. Catal., 7, 369 (1983). Such references also include, for example, M. S. Wainwright, xe2x80x9cRaney Copper and Raney Copper-Zinc Catalysts,xe2x80x9d Chem. Ind. (Dekker), 68, 213-30 (1996).
References describing the preparation of nickel/iron sponges include, for example, H. J. Becker and W. Schmidt in xe2x80x9cRaney nickel-iron catalyst,xe2x80x9d Ger. Offen. DE 2713374 19780928 (1978).
References describing the preparation of nickel/cobalt sponges include, for example, J. P. Orchard, A. D. Tomsett, M. S. Wainwright, and D. J. Young in xe2x80x9cPreparation and Properties of Raney Nickel-Cobalt Catalysts,xe2x80x9d J. Catal., vol. 84, pp. 189-99 (1983).
Various metal sponges are also commercially available from, for example, W. R. Grace and Co. (Chattanooga, Tenn.); Gorwara Chemical Industries (Udaipur, India); Activated Metals and Chemicals, Inc. (Sevierville, Tenn.); Degussa Huls Corp. (Ridgefield Park, N.J.); Engelhard Corp. (Iselin, N.J.); and Aldrich Chemical Co. (Milwaukee, Wis.).
Examples of suitable commercially-available nickel sponges, for example, include Raney(copyright) 2800 (characterized by the manufacturer as having at least 89 wt. % Ni; no greater than 9.5 wt. % Al; no greater than 0.8 wt. % Fe; an average particle size in the range of 20-60 xcexcm; a specific gravity of approximately 7; and a bulk density of 15-17 lbs/gal based on a catalyst slurry weight of 56% solids in water), Raney(copyright) 4200 (characterized by the manufacturer as having at least 93 wt. % Ni; no greater than 6.5 wt. % Al; no greater than 0.8 wt. % Fe; an average particle size in the range of 20-50 xcexcm; a specific gravity of approximately 7; and a bulk density of 15-17 lbs/gal based on a catalyst slurry weight of 56% solids in water), Raney(copyright) 4310 (characterized by the manufacturer as having at least 90 wt. % Ni; no greater than 8 wt. % Al; 0.5-2.5 wt. % Mo; no greater than 0.8 wt. % Fe; an average particle size in the range of 20-50 xcexcm; a specific gravity of approximately 7; and a bulk density of 15-17 lbs/gal based on a catalyst slurry weight of 56% solids in water), Raney(copyright) 3110 (characterized by the manufacturer as having at least 90 wt. % Ni; 0.5-1.5 wt. % Mo; no greater than 8.0 wt. % Al; no greater than 0.8 wt. % Fe; an average particle size in the range of 25-65 xcexcm; a specific gravity of approximately 7; and a bulk density of 15-17 lbs/gal based on a catalyst slurry weight of 56% solids in water), Raney(copyright) 3201 (characterized by the manufacturer as having at least 92 wt. % Ni; no greater than 6 wt. % Al; no greater than 0.8 wt. % Fe; 0.5-1.5 wt. % Mo; an average particle size in the range of 20-55 xcexcm; a specific gravity of approximately 7; and a bulk density of 15-17 lbs/gal based on a catalyst slurry weight of 56% solids in water), Raney(copyright) 3300 (characterized in U.S. Pat. No. 5,922,921 as having 90-99.1 wt. % Ni; no greater than 8.0 wt. % Al; no greater than 0.8 wt. % Fe; 0.5-1.5 wt. % Mo; no greater than 0.8 wt. % Ni; an average particle size in the range of 25-65 xcexcm; a specific gravity of approximately 7; and a bulk density of 15-17 lbs/gal based on a catalyst slurry weight of 56% solids in water), Raney(copyright) 2724 (Cr-promoted), and Raney(copyright) 2724 (Cr-promoted), all sold by W. R. Grace and Co.; the catalyst described as xe2x80x9cRaney nickelxe2x80x9d sold by Gorwara Chemical Industries; A-4000 and A-5000, sold by Activated Metals and Chemicals, Inc.; nickel ABMC, sold by Degussa Huls Corp.; and xe2x80x9cRaney nickel,xe2x80x9d Catalog No. 22,167-8, sold by Aldrich Chemical Co.
Examples of suitable commercially-available cobalt sponges include Raney(copyright) 2700 (characterized in U.S. Pat. No. 5,922,921 as having 93.0 wt. % Co; no greater than 6.0 wt. % Al; no greater than 0.7 wt. % Fe; no greater than 0.8 wt. % Ni; an average particle size in the range of 20-50 xcexcm; a specific gravity of approximately 7; and a bulk density of 15-17 lbs/gal based on a catalyst slurry weight of 56% solids in water), sold by W. R. Grace and Co.; the cobalt sponge catalysts purportedly manufactured by the Raney process and sold by Activated Metals and Chemicals, Inc.; and cobalt ABMC, sold by Degussa Huls Corp.
b. Deposition of the Copper Coating
Copper (typically in the form of a copper coating) may be deposited onto the surface of a metal support using various techniques well-known in the art for depositing metal onto metal surfaces. These techniques include, for example, liquid phase methods, such as electrochemical displacement deposition and electroless plating; and vapor phase methods such as physical deposition and chemical deposition. The following discussion will focus on the two particularly preferred techniques: electrochemical displacement deposition and electroless plating. This preference stems from the fact that the other techniques are generally more complicated and/or more costly.
i. Electrochemical Displacement Deposition of Copper
The copper coating may be deposited onto the support surface via electrochemical displacement deposition wherein copper ions in a copper-salt solution in contact with the support are reduced to copper metal as non-copper metal near the surface of the support is oxidized. The copper metal, in turn, forms a coating on the surface of the support, while the non-copper ions go into solution. A general discussion related to electrochemical displacement deposition may be found in, for example, G. A. Krulik and N. V. Mandich, xe2x80x9cMetallic Coatings (Survey)xe2x80x9d, Kirk-Othmer Encyclopedia of Chemical Technology, 4th Ed. (J. I. Kroschwitz and M. Howe-Grant, eds., Wiley, New York, N.Y., 1995) Vol. 16, pp. 258-91.
Suitable copper salts for displacement deposition include, for example, the nitrate, sulfate, chloride, acetate, oxalate, and formate salts of copper (this list is not exhaustive). Salts containing copper in the divalent state (i.e., Cu(II)) are typically the most preferred (although salts containing monovalent and trivalent copper may be used, they are typically less preferred because they tend to be unstable, commercially less available, and/or insoluble in the alkaline mixture).
Before and during the displacement deposition, the metal support preferably is protected from air by, for example, keeping it immersed in water, maintaining it under a non-oxidizing atmosphere (noble gas or N2, preferably N2), and/or sparging a suspension containing the support with a non-oxidizing gas. In one particularly preferred embodiment, the metal support surface is reduced before the displacement deposition. The surface may be reduced by, for example, contacting it with a solution of sodium borohydride (NaBH4), formaldehyde, or other reducing agent; or by contacting it with H2 or another reducing gas at an elevated temperature. Example 5 demonstrates such a technique.
To initiate the displacement deposition, the copper salt may be added as a dry powder to a solution containing the metal support, but more preferably is added as an aqueous solution. While adding the copper salt, the solution containing the metal support preferably is gently stirred at a rate to keep the support particles suspended. Although the copper salt may be added all at once, it is preferable to add the salt slowly so that the salt concentration does not exceed the concentration at which the salt begins to precipitate. Typically, the salt is added over a period of at least about 30 minutes, but no greater than about 2 hours (such slow salt addition is often unnecessary in the presence of a strong chelating agent, such as ethylenediaminetetraacetic acid, which keeps the copper salt solubilized). After the salt has been added, the resulting mixture preferably is stirred for at least about 15 minutes. Afterward, the stirring may be discontinued so that the catalyst can settle to allow the supernatant to be removed by decantation or other means. The catalyst may then be re-suspended in the desired solvent for introduction into the dehydrogenation reaction zone.
During the displacement deposition, the pH of the solution containing the metal support preferably is adjusted so that the displaced metal will tend to remain soluble and not redeposit onto the support. Metal ions are generally more soluble under acidic conditions than basic conditions (with the exception of alkali metal ions, which are generally soluble under both acidic and basic conditions). Thus, the pH is preferably low enough to ensure that the displaced metal remains in solution and does not redeposit onto the catalyst as, for example, an oxide or hydroxide.
If, during the displacement deposition, the copper is deposited at a rate which tends to unevenly coat the support, a more even coating may often be obtained by including a protecting chelating agent in the copper salt solution to control (i.e., slow) the rate of copper deposition so that a more even coat may be obtained. A chelating agent may also be beneficial to inhibit the displaced metal from redepositing onto the metal support. Often suitable chelating agents include, for example, hydroxy carboxylic acids (e.g., lactic acid, malic acid, citric acid, and tartaric acid) and salts thereof (e.g., sodium potassium tartrate, also described in the art as xe2x80x9cRochelle saltxe2x80x9d), with tartaric acid and salts thereof being particularly preferred. Chelators which contain amines (e.g., salts of iminodiacetic acid, nitrilotriacetic acid, and particularly ethylenediaminetetraacetic acid (also known as xe2x80x9cEDTAxe2x80x9d)) inhibit plating, and are therefore generally less preferred. Normally, at least one molar equivalent (based on moles of copper ions) of chelating agent is preferably included. Even more preferably, from about 1.2 to about 3.0 (still even more preferably from about 1.2 to about 1.8) molar equivalents of chelating agent are included in the mixture. Although concentrations of greater than 3.0 molar equivalents may be used, such additional concentrations usually do not provide any greater benefits. Concentrations of greater than 3.0 molar equivalents also tend to cause the chelating agent to precipitate and may create greater burdens downstream during product purification.
Examples 1, 3, 5, and 7 illustrate electrochemical displacement deposition of copper onto a metal sponge support. Those examples also illustrate the use of a chelating agent during such a deposition.
ii. Electroless Plating of Copper
Electroless plating may alternatively be used to deposit copper onto the surface of the support. Like displacement deposition, electroless plating comprises reducing copper ions to copper metal in a solution in contact with the support. However, unlike displacement deposition, substantially all the copper ions are reduced by an external reducing agent rather than the support itself. As the reducing agent reduces the copper ions in the solution to copper metal, the copper metal forms a coating on the surface of the support. It is generally preferred for electrochemical displacement plating to be suppressed during electroless plating. This is preferably accomplished by the presence of chelators, such as the amine chelators discussed above (particularly salts of EDTA). The chelator is preferably added to the copper salt or the metal support before the copper salt and metal support are mixed to avoid electrochemical displacement plating from occurring in the absence of the reducing agent.
A wide variety of suitable reducing agents may be used. These include, for example, sodium hypophosphite (NaH2PO2), formaldehyde (CH2O) and other aldehydes, formic acid (HCOOH), salts of formic acid, salts of borohydride (e.g., sodium borohydride (NaBH4), salts of substituted borohydrides (e.g., sodium triacetoxyborohydride (Na(CH3CO2)3BH), sodium alkoxides, and hydrazine (H2NNH2). Sodium borohydride is particularly preferred because it is readily available, may be solubilized without heating, and has sufficient activity at room temperature to enable plating to be completed within about 1 hour.
Typically, the reducing agent is added slowly (preferably over a period of from about 5 minutes to 3 hours, and more preferably from about 15 minutes to about 1 hour) to a slurry of the metal support in water or an alcohol under an inert atmosphere (e.g., N2). If the reducing agent is instead first added to the copper salt, it is preferably added to a solution which contains the copper salt and also a chelator (the presence of the chelator inhibits the reduction of the copper ions before the copper-salt solution is combined with the metal support).
The metal support preferably is essentially free of surface oxidation at the time of the plating. Consequently, in instances where the metal support has an oxidized surface (such as when the support has been exposed to air (even while under water) for 6 or more months), it is particularly preferable to pre-treat the support with a reducing agent. Here, for example, the support may be stirred in a sodium borohydride solution (preferably a solution having a pH of at least about 10 and comprising at least about 1 gram of sodium borohydride per 25 grams of metal support) for from about 15 minutes to about 2 hours at room temperature.
When the metal support comprises at least about 50 wt. % nickel, it is typically more preferred to deposit the copper coating using electroless plating rather than electrochemical displacement plating. This preference stems from the fact that electroless plating of copper onto such metal supports often tends to produce catalysts which have superior activity and selectivity, particularly when the catalyst is used to catalyze the dehydrogenation of diethanolamine to form the salt of iminodiacetic acid.
Examples 9, 11, and 13 illustrate the use of electroless plating to deposit copper onto the surface of a metal support.
2. Unsupported Copper-containing Catalysts
In another embodiment of this invention, the catalyst does not comprise copper coated on a metal support (i.e., there is no discrete copper coating forming the surface of the catalyst). Rather, the copper is simply mixed (preferably in the form of an alloy) with other metals which provide desirable properties. In this embodiment, from about 10 to about 85% (more preferably from about 50 to about 85%, and even more preferably from about 60 to about 80%, and still more preferably from about 60 to about 75%) by weight of the catalyst is copper. Preferably, the catalyst is in the form of a metal sponge. In a particularly preferred embodiment, the catalyst comprises greater than about 1% (by weight) nickel, tin, or a combination thereof. In another particularly preferred embodiment, the catalyst comprises less than about 1% (by weight) metal oxide.
It should be recognized that this embodiment is less preferred if there are significant adverse effects from the non-copper metal of the catalyst being in contact with the other components in the reaction zone. For example, a catalyst having a copper coating is more preferred if the catalyst contains a metal which catalyzes an undesirable side reaction that reduces the conversion of the primary alcohol and/or selectivity for the desired carboxylic acid salt. This occurs, for example, where a catalyst containing nickel is used to dehydrogenate diethanolamine to form a salt of iminodiacetic acid. Without a copper coating, the exposed nickel tends to catalyze the formation of a glycine salt byproduct, which, in turn, reduces the selectivity for the desired iminodiacetic acid salt. By using a copper coating, however, the activity of the nickel can often be minimized.
A copper coating is also preferred if, for example, a non-copper metal in the catalyst is vulnerable to attack under the reaction conditions to an extent which significantly reduces the lifetime of the catalyst absent a copper-containing coating. Metals which are often vulnerable to such attack include zinc, tin, cobalt, and iron.
3. Optional Modifier Metal
The catalyst may optionally contain one or more supplemental metals (i.e., modifier metals) selected from the group consisting of chromium, titanium, niobium, tantalum, zirconium, vanadium, molybdenum, manganese, tungsten, bismuth, tin, antimony, lead, and germanium. The presence of such a metal(s) tends to extend the life of the catalyst, i.e., increase the number of reaction runs in which the catalyst can be used before its activity decreases to unacceptable levels. Of the above elements, vanadium, chromium, molybdenum, and mixtures of chromium and molybdenum are particularly preferred.
The amount of the modifier metal(s) can vary within wide limits. Preferably, the total concentration of modifier metals is at least about 10 parts per million parts of copper in the catalyst by weight. More preferably, the total concentration of the modifier metals in the catalyst is from about 0.002 and to about 5% by weight, more preferably from about 0.002 to about 2.5% by weight, even more preferably from about 0.005 to about 2% by weight, and still even more preferably from about 0.5 to about 1.5% by weight. Typically, the total concentration of modifier metals does not exceed about 5% by weight. Although greater concentrations of modifier metals can be used, no additional benefits are usually obtained by exceeding such a concentration and the activity of the catalyst is generally reduced.
The modifier metal(s) may be contained in the metal support and/or copper coating. Where it is desirable to include the modifier metal(s) in an alloy-metal support, the modifier metal(s) are preferably incorporated into the alloy at the time the alloy is formed. Where it is desirable to include the modifier metal(s) in the copper coating, the modifier metal may, in some instances, be deposited simultaneously with the copper. Where, however, the copper is deposited via displacement deposition or electroless plating (discussed above), the modifier metal(s) are preferably added to the catalyst after the copper has been deposited because the modifier metals tend to dissolve under displacement deposition conditions and to inhibit electroless plating. A modifier metal(s) may typically be added to the catalyst surface by simply contacting the catalyst with an aqueous solution containing a salt (e.g., a sulfate, nitrate, chloride, etc.) of the modifier metal(s).
C. Preferred Reaction Conditions
This dehydrogenation reaction is conducted in an alkaline environment (i.e., a basic environment). More specifically, this reaction is typically conducted in the presence of a strong base having a pKa value of at least about 11, more preferably at least about 12, and even more preferably at least about 13. Suitable bases include, for example, alkali metal hydroxides (LiOH, NaOH, KOH, RbOH, or CsOH), alkaline-earth metal hydroxides (e.g., Mg(OH)2 or Ca(OH)2), NaH, and tetramethyl ammonium hydroxide. Of these bases, alkali metal hydroxides (particularly NaOH and KOH, and even more particularly NaOH) are often preferred because of their solubility in water under the reaction conditions, as well as their ready commercial availability and ease of handling.
The preferred amount of base introduced into the reaction zone depends on, for example, the moles of primary alcohol groups introduced into the reaction zone. Preferably, at least about one molar equivalent of base is introduced per mole of primary alcohol hydroxy groups. Thus, for example, if the base is NaOH and the primary alcohol is monoethanolamine, preferably at least about 1 mole of NaOH is introduced per mole of monoethanolamine. If, on the other hand, the primary alcohol is diethanolamine, preferably at least 2 moles of NaOH are introduced per mole of diethanolamine. In a particularly preferred embodiment, from about 1.05 to about 2.0 molar equivalents of base per alcohol hydroxyl group are introduced. The hydroxide may, for example, be in the form of flakes, powder, pellets, or an aqueous solution.
The reaction is normally conducted in a solvent in which the base is soluble. Preferably, a sufficient quantity of solvent is present in the reaction zone to dissolve essentially all (more preferably, all) the base. The solvent also preferably is present in a sufficient quantity to maintain the primary alcohol substrate and carboxylic acid salt product in a solubilized form. Water is normally the preferred solvent due to its low cost, widespread availability, and ease of handling.
The preferred catalyst loading (i.e., the preferred amount of catalyst introduced into the reaction zone) depends on, for example, the amount of the primary alcohol substrate introduced into the reaction zone. Typically, the catalyst loading is at least about 1% by weight of the primary alcohol substrate (i.e., [mass of catalyst÷mass of primary alcohol substrate]xc3x97100%). More preferably, the catalyst loading is from about 1 to about 70% (still more preferably from about 10 to about 40%) by weight of the primary alcohol substrate.
The preferred catalyst loading also depends on, for example, the amount of total reaction mass. Typically, the catalyst loading is at least about 0.1% by weight of the total reaction mass (i.e., [mass of catalyst÷total reaction mass]xc3x97100%). More preferably, the catalyst loading is from about 0.1 to about 10% (even more preferably from about 3.5 to about 10%, and still even more preferably from about 3.5 to about 5%) by weight of the total reaction mass. Concentrations of greater than about 10 wt. % are difficult to filter. On the other hand, concentrations of less than about 0.1 wt. % tend to produce unacceptably low reaction rates.
It has been found that the catalysts of this invention typically are able to achieve a greater activity than the same volume of traditional copper-on-carbon catalysts (i.e., the catalysts of this invention typically have a greater activity per unit volume relative to the traditional copper-on-carbon catalysts). This greater activity per unit volume is advantageous because it often makes such catalysts easier to filter, thereby increasing throughput. Without being bound by any particular theory, Applicants believe that the greater activity per unit volume may be due, at least in part, to the catalysts of the present invention having a greater copper surface area relative to traditional copper-on-carbon catalysts.
The reaction typically is conducted at a temperature of at least about 70xc2x0 C., preferably from about 120 to about 220xc2x0 C., more preferably from about 140xc2x0 C. to about 200xc2x0 C., even more preferably from about 145 to about 155xc2x0 C., and still even more preferably about 150xc2x0 C. (particularly when the primary alcohol is diethanolamine and the desired product is the salt of iminodiacetic acid). Although a reaction temperatures outside of these ranges may be used, the results are typically less than optimal. For example, at temperatures of less than about 120xc2x0 C., the reaction rate tends to be slow. And at temperatures greater than about 220xc2x0 C., the catalyst normally begins to lose selectivity. To illustrate, as the reaction temperature exceeds about 150xc2x0 C. (and particularly as the temperature exceeds about 220xc2x0 C.), the dehydrogenation reaction of diethanolamine will tend to form more glycine salt byproduct, and, therefore, be less selective toward forming the desired iminodiacetic acid salt product.
The reaction is preferably conducted under pressure. More specifically, the reaction is normally conducted under a pressure which is sufficient to prevent boiling of the mixture at the reaction temperature. At reaction temperatures of from about 120 to about 220xc2x0 C., the pressure preferably is at least about 5 kg/cm2, more preferably from about 5 to about 30 kg/cm2, even more preferably from about 5 to about 20 kg/cm2, still even more preferably from about 8 to about 11 kg/cm2 (i.e., from about 115 to about 155 psig), and most preferably about 9.4 kg/cm2 (i.e., 135 psig). Although greater pressures may be used, they are normally less desirable because they tend to reduce the reaction rate.
The dehydrogenation reaction preferably is conducted under a non-oxidizing atmosphere (preferably, an atmosphere containing a noble gas and/or N2, and more preferably N2 when the reaction is conducted on a commercial level) to avoid oxidation of the catalyst surface (the atmosphere will also contain H2 which evolves during the dehydrogenation). This preference stems from the fact that oxidization of the copper near the surface of the catalyst tends to reduce the activity and selectivity of the catalyst.
The dehydrogenation reaction may be carried out in a wide variety of batch, semi-batch, and continuous reactor systems. The configuration of the reactor is not critical. Suitable conventional reactor configurations include, for example, stirred-tank reactors, fixed bed reactors, trickle bed reactors, fluidized bed reactors, bubble flow reactors, plug flow reactors, and parallel flow reactors. Often, the more preferred reactor configurations are stirred-tank reactors.
When the dehydrogenation is conducted in a continuous reactor system, the residence time in the reaction zone can vary widely depending on the specific catalyst and conditions employed. Likewise, when the dehydrogenation is conducted in a batch reactor, the reaction time typically will also vary widely depending on such factors. Normally, the dehydrogenation behaves as a first order reaction, particularly toward the end of the reaction. Thus, the preferred residence time in a continuous reaction zone (or the preferred reaction time in batch reaction zone) will also depend on the desired degree of conversion.
D. Use of Carboxylic Acid Salt Product to Make N-(phosphonomethyl)glycine or a Salt Thereof
Various carboxylic acid amine salts produced by this invention may be used as raw materials to prepare N-(phosphonomethyl)glycine and agronomically acceptable salts thereof in accordance with many well-known methods in the art. As used herein, an xe2x80x9cagronomically acceptable saltxe2x80x9d is defined as a salt which contains a cation(s) that allows agriculturally and economically useful herbicidal activity of an N-(phosphonomethyl)glycine anion. Such a cation may be, for example, an alkali metal cation (e.g., a Na ion), an ammonium ion, an isopropyl ammonium ion, a tetra-alkylammonium ion, a trialkyl sulfonium ion, a protonated primary amine, a protonated secondary amine, or a protonated tertiary amine.
Particularly preferable carboxylic acid amine salts that may be produced by this invention and used for making N-(phosphonomethyl)glycine compounds are salts of iminodiacetic acid (particularly alkali metal salts of iminodiacetic acid). These carboxylic acid salts may be phosphonomethylated in a reaction zone containing HCl, phosphorous acid (H3PO3), and formaldehyde (CH2O) to form N-(phosphonomethyl)iminodiacetic acid. See, e.g., Gentilcore, U.S. Pat. No. 4,775,498 (also reporting that the HCl and H3PO3 may optionally be formed by adding PCl3 to water). The N-(phosphonomethyl)iminodiacetic acid may, in turn, be contacted with oxygen in the presence of a catalyst to oxidatively cleave a carboxymethyl group to form N-(phosphonomethyl)glycine. Many catalysts are known in the art for conducting this oxidation, and include, for example, carbon catalysts (see, e.g., Hershman, U.S. Pat. No. 3,969,398; and Chou, U.S. Pat. Nos. 4,624,937 and 4,696,772); a carbon catalyst along with a noble metal co-catalyst supported on aluminosilicate (see, e.g., Felthouse, U.S. Pat. No. 4,582,650), and catalysts comprising a noble metal supported on carbon (see, e.g., Franz, U.S. Pat. No. 3,950,402; Ramon et al., U.S. Pat. No. 5,179,228; and Ebner et al., PCT/US99/03402).
Alternatively, for example, a salt of glycine (particularly an alkali metal salt of glycine) may be converted to N-phosphonomethyl)glycine by a wide variety of methods well-known in the art. Many such methods are summarized in Franz, et al., Glyphosate: A Unique Global Herbicide (ACS Monograph 189, 1997) at pp. 234-39.
As a further example, a salt of an N-substituted glycine (e.g., a salt of N-(methyl)glycine, also known as xe2x80x9csarcosinexe2x80x9d) may be phosphonomethylated by, for example, reacting it with PCl3 in water, and then filtering out the resulting salt and adding CH2O. The resulting product is an N-substituted-N-(phosphonomethyl)glycine (e.g., N-methyl-N-(phosphonomethyl)glycine). A solution containing the N-substituted-N-(phosphonomethyl)glycine) may then be contacted with oxygen in the presence of a noble metal catalyst (preferably platinum) to form N-(phosphonomethyl)glycine. See Morgenstern et al., U.S. Pat. No. 6,005,140. Other approaches for making N-(phosphonomethyl)glycine from N-substituted glycine compounds include, for example phosphonomethylating N-benzylglycine to form N-benzyl N-(phosphonomethyl)glycine, and then (a) reacting the N-benzyl N-(phosphonomethyl)glycine with hydrobromic or hydroiodic acid to cleave the benzyl group (see, e.g., Parry et al., U.S. Pat. No. 3,956,370), or (b) converting the N-benzyl N-(phosphonomethyl)glycine to N-(phosphonomethyl)glycine via hydrogenolysis (see, e.g., European Patent Application No. 55,695; and Maier, L., Phosphorus, Sulfur and Silicon, 61, 65-7 (1991)); and phosphonomethylating N-t-butylglycine to form N-t-butyl N-(phosphonomethyl)glycine, and then converting the N-t-butyl N-(phosphonomethyl)glycine to N-(phosphonomethyl)glycine via acid hydrolysis (see Gaertner, U.S. Pat. No. 3,927,080).